Free Meta Ads report template (May 2026 edition)

This Meta Ads report template is the one to use in May 2026 — built around the metrics that still exist after this year's attribution overhaul. The 7-day-view and 28-day-view columns are gone. Click-through now counts only link clicks. A new "engage-through" bucket holds likes, shares, and saves on a 1-day window. If your weekly template hasn't been rewritten this year, it's measuring two different things and comparing them as if they're the same.

This is the structure to copy. Skip to the action list section if you only want the part that drives Monday decisions.

Why every 2025 template is now broken

Three changes between January and March 2026 made the older Meta Ads report templates inaccurate:

1. Attribution windows shrank. According to Meta's Business Help Center, the default attribution window is now 7-day click + 1-day engage-through + 1-day view. The 7-day-view and 28-day-view columns were removed from the Ads Insights API in January 2026, as confirmed by DataSlayer's reporting on the change.
2. Click-through was redefined. As Jon Loomer breaks down, social interactions — likes, shares, saves — moved out of click-through and into the new engage-through bucket in March 2026. Same conversions, different column.
3. Manus AI now generates auto-summaries inside Ads Manager. Meta acquired Manus AI for $2B and rolled it out to advertisers on February 17, 2026 (Search Engine Land, Social Media Today, MediaPost). It produces report-style summaries on demand, which means the template you write now needs to be the layer Manus can't produce on its own.

A template designed before any of those changes will overstate conversions, double-count engaged-views as clicks, and skip the diagnosis layer that Manus doesn't auto-fill.

What goes in the May 2026 template (top to bottom)

Six sections. Action items, not analysis. The reader should be able to scan it in 90 seconds and know exactly what to do this week.

1. Headline diagnosis (one paragraph)

A plain-English summary of the week's state. Three sentences max. State the verdict, the cause, and the action. Example: "Account is on watch. Spend held flat at $24K, but ROAS dropped from 3.6 to 2.7 driven by the Spring-3 creative set hitting frequency 4.2 on cold. Two new creatives shipped Friday — rotate the tired set out by Wednesday."

If you can't write this paragraph in three sentences, you don't understand the week yet — go back to the data.

2. Executive scorecard (7-metric table, with the new windows)

The metric list hasn't changed, but the underlying numbers come from a different attribution window now. Make it explicit.

| Metric | Window in 2026 | What it tells you | Trigger for action | |---|---|---|---| | Spend | N/A | Budget pacing vs. plan | ±10% WoW outside intended change | | Conversions (link click) | 7-day click | Output from intended traffic | Down WoW while spend is flat or up | | Conversions (engage-through) | 1-day engage | Soft-touch incrementality signal | Track separately — don't blend with click | | ROAS | 7-day click only | Efficiency on the strict window | Outside account band by 15%+ | | CPA | 7-day click only | Unit economics | Above CAC ceiling 3+ days | | CTR (link) | N/A | Creative health | Below account median 7+ days | | Frequency | 7-day rolling | Saturation | >2.5 on cold (tightened in 2026) | | CVR (LP → purchase) | 7-day click | LP + offer health | Drop >20% WoW with no LP change |

Two notes that almost every template gets wrong:

• Don't sum click-through and engage-through. They're different conversion qualities. Per the Meta help docs, engage-through includes likes, shares, saves, and engaged-views on a 1-day window — fundamentally a different intent signal than a link click.
• Annotate the year-over-year comparison. Any May-2026-vs-May-2025 row is comparing two measurement systems. Flag it explicitly or drop it.

3. Account Health Score (composite)

A single 0–100 number that rolls up ROAS trend, frequency health, spend efficiency, and creative velocity. Without a composite, every metric in the scorecard reads as equally important and the reader has to do the synthesis themselves — which is exactly the work a good report is supposed to skip. See the GMC Meta Ads account health score for one version of the rollup.

4. Creative performance (with 2026 frequency thresholds)

Top 5 active creatives by spend, link CTR, and CVR. Plus a creative fatigue heatmap by creative age and frequency. The textbook fatigue signal in 2026 is frequency >2.5 on cold + declining link CTR over a 5–7 day window — tighter than the >3.5 threshold most 2024–2025 templates still use, because Advantage+ broader audiences cycle faster.

For the diagnostic side, see the GMC creative fatigue tool and the practitioner-level breakdown of Motion vs Triple Whale vs Good Morning on creative fatigue diagnosis.

5. Action list (tiered by urgency)

Three tiers. No more than 5 items total. Skip "Monitor" if there's nothing genuinely passive to track — padding the list dilutes the urgent items.

• Act today — bleeds money if it waits (kill a fatigued creative, pause a duplicated campaign, fix a broken pixel event).
• This week — moves the number by Friday (ship 2 new creatives, restructure a campaign, raise/lower budget on a winner).
• Monitor — watch but don't touch (a frequency edging toward 2.5, a soft CTR trend that hasn't broken yet).

Without tiering, every recommendation gets equal weight and nothing gets done by Friday.

6. What we did last week

Bullet log of actions taken since the prior report. Five items max. Critical for two reasons: it shows the account is being managed, not just observed, and it builds a retrospective trail when a creative or audience eventually breaks. For agency context specifically, see the breakdown of best Meta Ads reporting software for agencies.

What to cut from older templates

If your template still includes any of these, remove them or move them to a separate appendix:

• 7-day-view and 28-day-view columns. Removed in January 2026. Any cell pulling from those windows is now empty or stale.
• Combined click-through totals. Older templates summed clicks + engagements as a single conversion column. After the March 2026 redefinition that overstates click-through performance.
• Static year-over-year tables without an asterisk. YoY comparisons across the January 2026 break compare two different measurement systems. Annotate or skip.
• Audience breakdowns by placement on the front page. Useful for debugging — not for weekly decisions. Move to a deep-dive doc.
• Screenshots of Ads Manager. Nobody re-parses them. Pull the numbers into the report itself.

Manus AI changed the bar for what a report has to do

Per Adage, Manus inside Ads Manager can produce a report-style summary of any account on demand. That means the floor for a useful weekly report is no longer "list of metrics with deltas" — Manus does that for free. The differentiated layer is the diagnosis paragraph plus the urgency-tiered action list. Everything Manus can't auto-produce is where the human (or the next-gen reporting tool) earns its keep.

This is the same structural distinction operators have been making for a decade: dashboards describe, reports recommend. The bar just moved up because the description layer got automated.

Common mistakes

1. Pulling the scorecard from a saved view that still includes 7-day view. Many older saved views silently default to the removed window and now return null — re-create them on the new default.
2. Treating engage-through as a click. Different conversion quality, different prospect intent, different downstream LTV. Track it. Don't blend it.
3. Writing the report on Monday morning. Run the diagnostic by Sunday EOD. By Monday the action items have already burned a day.
4. Skipping the diagnosis paragraph because Manus produces one. Manus produces a competent description; the differentiated value is the opinion — what to do, why, by when.
5. Same template for every stakeholder. Founders want the paragraph + action list. Ops leads want the scorecard. Media buyers want the creative detail. Either produce three views or pick the most senior reader and optimize for them.

FAQ

What's the right default attribution window for the 2026 weekly report? 7-day click + 1-day engage-through + 1-day view, with click and engage tracked as separate columns. That matches Meta's current default attribution settings per the Business Help Center. Reporting on a non-default window is fine if you justify it in the headline, but mixing two different windows across rows of the same scorecard is not.

Should I keep showing year-over-year in May 2026? Only with an explicit annotation that the January 2026 attribution change broke the comparison. A YoY row without that note will be misread.

Does Manus AI replace the weekly report? No. Manus produces description-layer summaries on demand. The weekly report's job is the diagnosis paragraph and the urgency-tiered action list — neither of which Manus auto-produces in a way an operator can act on without re-doing the synthesis.

How long should the May 2026 weekly report be? Two pages. One for paragraph + scorecard + Health Score. One for creative + action list + last-week log. Anything longer is appendix.

What does this template look like compared to a deeper account audit? A weekly report is operational; an audit is structural. If the underlying account hasn't been reviewed in a quarter, run the 30-minute Meta Ads audit first and update the report template against what you find. For the standalone tool, see the Meta Ads audit tool.

Where do the 2026 attribution changes affect the report most? The scorecard. The action list mostly stays the same — fatigued creatives are still fatigued, wasted spend is still wasted spend. But every conversion-based row in the scorecard is now drawing from a different window than it was last year. For the deeper read, see the Meta attribution changes 2026 explainer.

CTA

Good Morning skips the "build the template" step entirely. The dashboard already runs the May 2026 metric set, separates click-through from engage-through, applies the tightened 2026 frequency thresholds, and outputs an urgency-tiered action list. No template to maintain, no attribution-window cleanup, no week-after-week formatting. See the action list in action →